Liu Qiaoyun messaged through the secure channel on Friday morning:
*The monitoring architecture is fully removed. I've confirmed the deployment mechanism shows no trace of the access layer. I sent you a verification signature through the system channel — your instruments should confirm it.* A pause in the message thread. *I also have something that may be useful to you. Xu Mingzhi told me about the Bureau researcher's work on the Awakening Event. I've read the classification archive materials she's been trying to access. All of them. I can deliver the content.*
He forwarded this to Han Weiwei immediately.
Her reply came in eight minutes: *When can she come in.*
---
Liu Qiaoyun arrived at the Bureau's fourteenth floor at noon. She looked at the field characterization instruments with the attention of someone who understood what she was seeing — not the technical specifications, but the function.
She set a worn leather portfolio on Han Weiwei's conference table and sat down.
"I've been in the classification archive twice," she said. "The first time was six years ago — I was still with the Bureau and had cleared access. I read everything that was available about the Awakening Event's immediate aftermath. The second time was eighteen months ago, using the hardware contact access my ability provides." She looked at Han Weiwei. "The materials you can't access through standard channels — I read them in physical contact with the archive's storage media. I have complete recall of their content."
Han Weiwei looked at her steadily. "Your ability provides recall, not reproduction."
"I can dictate what I read," Liu Qiaoyun said. "My recall is essentially photographic for documented material I've had direct contact with. I can reproduce the content verbally or in writing with high accuracy."
"How high?" Han Weiwei said.
"The kind of accuracy that comes from having read something fourteen times because you were certain it was important," Liu Qiaoyun said.
Han Weiwei opened a new document file.
"Start with the immediate post-event ability classification reports," she said. "Specifically, the records of ability emergence distribution across the primary study cities."
Liu Qiaoyun opened her portfolio. A handwritten index. She'd come prepared.
Chen sat at the calibration position and listened as Liu Qiaoyun began dictating.
The classification reports were, as he'd suspected, significantly more detailed than the public record. The population-level ability emergence breakdown — not just city-wide percentages but distribution by pre-existing cultivation background, documented ability precursor traits, correlation with specific cultivation schools. Han Weiwei typed, cross-referenced against her existing data, occasionally stopped Liu Qiaoyun to ask clarifying questions about methodology or notation.
They worked for two hours.
At the end of the two hours, Han Weiwei had the complete dataset she'd been building toward for four years.
She sat for a moment with the finalized model on her primary monitor. The complete cross-city comparison. The ability emergence distributions. The cultivation school correlations. The pre-event practitioner records mapped against the post-event ability landscape.
"The pattern is complete," she said. "The Awakening Event selected and amplified existing cultivation potential. The 3% were not randomly chosen — they were the individuals in each city whose cultivation had developed farthest in directions compatible with the event's amplification mechanism."
"Yes," Liu Qiaoyun said.
"And in cities with strong probability cultivation traditions—"
"The probability-adjacent ability types predominate," Liu Qiaoyun said. "This city had a pre-event probability cultivation school with forty-three registered practitioners. Forty-one of them were in the 3% selection. The ability types they manifested were all probability-adjacent."
Chen looked at the data. Forty-three practitioners. Forty-one selected. Nearly the entire school.
"The event selected for cultivation quality, not raw potential," he said. "Advanced practitioners were amplified regardless of what school they were in."
"The two who weren't selected," Han Weiwei said. "Do you know why?"
"One had been inactive in cultivation practice for eight years due to illness," Liu Qiaoyun said. "The other was registered but had been practicing at a minimal level — administrative registry entry, not active cultivation." She paused. "The event required active engagement. Dormant potential wasn't enough."
He thought about this. The 97% — people with potential but not with active cultivation development far enough along. Not excluded by ability but by state.
"The compensatory protocol," he said. "For zero-assessments — people with no cultivation background at all — the event couldn't find anything to amplify. So the compensatory mechanism was triggered for specific cases."
"Not all zero-assessments," Liu Qiaoyun said. "The classification records show the compensatory protocol was selective. Forty-seven zero-assessments in this city, and only one received a compensatory system activation." She looked at him. "You."
He sat with this.
One. Forty-seven zero-assessments, one selected for compensatory protocol.
"The selection criteria," he said. "The records say anything about how the compensatory protocol chose?"
"The classification record for your case," she said, "notes: *Subject demonstrates unusually high baseline cognitive and analytical capacity combined with consistent applied problem-solving under high-stakes conditions over an extended period. Compensatory protocol activated for high-potential candidates whose cultivation background was insufficient for standard selection.* " A pause. "That's the full entry."
High baseline cognitive and analytical capacity. Applied problem-solving under high stakes over time.
Not luck. Not destiny.
He had been selected because he'd spent four years working two jobs to get through university while tutoring mathematics to anyone who needed it, solving problems under financial pressure that most of his cohort didn't have, thinking through every option before acting because there had never been a margin for error.
The event had looked at him and seen what the power world's hierarchy hadn't. Not potential to be a great cultivator. Potential to work a system.
Han Weiwei looked at the classification record Liu Qiaoyun had just dictated. She typed it, cross-referenced it against her model, ran the comparison.
"The compensatory protocol criteria," she said. "Cognitive and problem-solving capacity combined with demonstrated pressure performance. That's a different selection set than cultivation quality."
"It's selecting for different eventual utility," Liu Qiaoyun said.
Han Weiwei looked at Chen. "The architect's design — the compensatory protocol selects candidates who will be effective at operating the system, not candidates who had the right cultivation base to be amplified directly."
"Agents," he said.
"Someone to run the network," Liu Qiaoyun said.
The three of them sat with this.
The architect had designed the compensatory protocol to find people capable of coordinating the convergence architecture in Phase 3. Not the strongest cultivators. People who could work complex systems under pressure, who thought in terms of leverage and accumulated advantage, who would use the system rather than be used by it.
"The forty-six other zero-assessments," he said. "They didn't qualify."
"Their records show standard pressure performance or below," Liu Qiaoyun said. "The event assessed them and determined the compensatory threshold wasn't met."
He thought about the forty-six people who had received the same UNAWAKENED stamp and moved on with their lives. Some of them had probably been angry. Some had probably accepted it. None of them had received a system notification at eleven PM on a Tuesday night.
He had been selected. Specifically. Because he'd spent four years demonstrating what the event was looking for.
He noted this. Filed it under the category of things he'd known abstractly and now understood precisely.
---
After Liu Qiaoyun left, Han Weiwei ran the final model compilation for two hours.
He stayed. Not for the data — she had the data. For the process. Watching her build the complete picture from everything she'd been accumulating for four years, plus two hours of Liu Qiaoyun's dictated archive access, was something worth being present for.
At four-thirty, she saved the model and sat back.
"The paper," he said.
"A year of analysis before I publish," she said. "The methodology needs to be bulletproof before this goes to the academic record. The claim — that the Awakening Event was a designed selection mechanism — is extraordinary. The evidence standard has to match."
"You have the data."
"I have the data," she said. "I need the analytical framework that makes it impossible to dismiss as coincidence."
He thought about Xu Mingzhi's theoretical paper. The framework for population-level probability field deployment. "You need the theoretical mechanism."
"Xu Mingzhi's paper," she said.
"Yes."
She looked at him. "He should be a collaborator on this."
He sat with this. The sixty-three-year-old cultivation theorist who had built six years of monitoring architecture based on a genuine concern about network exploitation. His theoretical paper, the framework for the architect's design. Han Weiwei's empirical dataset, four years of building the evidence that the design was real.
Together, they'd have the theory and the evidence.
"I'll arrange it," he said.
She looked at the model on her monitor. Then back at him. The careful read — the version she did when she was thinking about something adjacent to the research.
"Liu Qiaoyun," she said. "She worked for Xu Mingzhi because she believed the network was exploitative. She had the archive data for six years and used it to build a counter-case."
"Yes," he said.
"And now she's sitting across from me helping me build the affirmative case," she said. "That's a significant change in orientation."
"She had the right concern about the wrong premise," he said. "When the premise changed, the orientation changed with it."
"She's good," Han Weiwei said. Not a compliment so much as an accurate assessment. "Her recall is extraordinary. The way she organized the archive content in her index — she'd been thinking about it for a long time."
"Six years," he said.
"She should be on the research collaboration," Han Weiwei said. "Not as a coauthor — as a data resource. Her ability to directly access archived documentation is something I'd use regularly if it were available."
He noted this. The network was expanding in directions he hadn't fully planned. Xu Mingzhi's theoretical framework. Liu Qiaoyun's archive access. Peng Lihua's empathic ability on the security committee.
The system had been building his operational environment. But the operational environment he was building was extending beyond what the system's original design accounted for.
Or the architect had accounted for it and hadn't told him yet.
He thought about the observer feed — Song Meiqi running at full output, the architect watching through her, seeing all of it.
"Song Meiqi," he said, thinking aloud.
Han Weiwei looked at him.
"The observer function," he said. "The architect's been watching through the observer feed since Song Meiqi joined. He's seen Xu Mingzhi, Liu Qiaoyun, the archive data, your research model. He's watching the network's expansion beyond his original seventeen-node design."
"And?" she said.
"And he's still not intervening," he said. "He's been patient for six years. He's patient now. But there's a threshold where watching ends and he acts."
"What's the threshold?" she said.
"Phase 2 completion," he said. "When the Level 5 threshold is crossed, Phase 3 activates — convergence. The observer's function changes. The network shifts from distributed operation to coordinated operation. The architect designed that transition point." He looked at the aggregate graph on her monitor. "When the network reaches that threshold, he has to engage directly. The convergence needs a coordinator."
"You," she said.
"The compensatory protocol selected me for it," he said. "Phase 3's coordinator function is why I was selected over forty-six other zero-assessments."
She sat with this. The model, the data, the complete picture they'd been building from different angles for months.
"When the architect contacts you," she said, "what are you going to say to him?"
He thought about the honest answer. "I'm going to ask him what Phase 3's shared objective is. The system's documentation says the nodes transition to coordinated operation toward a shared objective. It doesn't specify the objective."
"He knows," she said.
"He's known for longer than the network has existed," he said.
She looked at her monitor. The model. The probability field graph, the aggregate climbing toward its Phase 2 target output.
"The shared objective," she said. "Based on everything I've compiled about the Awakening Event's design — the selection mechanism, the seventeen nodes, the cultivated ability distribution—" She paused. "It's not city-level. Whatever the architect designed Phase 3 for, it operates at a scale above any single city."
He'd had the same thought. Seventeen nodes, distributed globally. Not one city. The whole network.
"The convergence architecture operates at global scale," he said.
"Which means when Phase 3 activates," she said, "the city's power hierarchy is not the relevant frame."
He looked at the probability field graph. The field running warm around him, the aggregate climbing.
Level 5 threshold. Still months away.
He had months to understand what the architect was building before it became relevant.
He noted this. Filed it.
"The analysis," he said. "Take the time you need. The paper needs to be right."
"It will be," she said.
---
Song Meiqi sent at eight PM:
*Full output all week. The aggregate is at ninety-three percent of the pre-quiet period peak. Five to seven days until we're back at full Phase 2 output.*
He read this and thought about the Level 5 timeline. At full Phase 2 output, the compressed timeline was six to eight months. Han Weiwei's calculation from several weeks ago.
The second party situation had cost them three weeks. The architecture was clean now. The nodes were running.
He sent back: *Good. Tell me when you hit full output.*
Her response: *You'll feel it before I tell you.*
He thought this was probably accurate.
